Evaluation of three methods for detecting low-level bacterial contamination in intravenous solutions.

The accuracy of three sterility-testing methods in detecting low-level contamination in deliberately contaminated intravenous solutions was studied. One-liter bags of 5% dextrose (D5W) and 0.9% sodium chloride (saline injections were contaminated with Staphylococcus epidermidis and Pseudomonas aeruginosa; approximately 10(1) viable bacteria were injected into each bag. Two membrane-filtration methods (Ivex-2 and Addi-Chek) and one aliquot method [twice concentrated trypic soy broth (2X-TSB)] were used to test each of 10 diliberately contaminated solutions for both D5W and saline; 500 ml of each liter bag was filtered or added to 2X-TSB. Incubation containers were stored at 37 degrees C and inspected at 24, 48, and 72 hours for turbidity. There was no significant difference among the three methods in the detection of contaminated saline solutions. The Addi-Chek system was significantly less effective in detecting contamination in D5W than either of the other methods. It is concluded that the Ivex-2 system is accurate and the easiest-to-use system of the three tested.